Cathodoluminescence Mapping for the Determination of N-Type Doping in Single Gaas Nanowires

Home , Cathodoluminescence

Cathodoluminescence mapping for the determination of n-type doping in single GaAs nanowires Hung-Ling Chen, Chalermchai Himwas, Andrea Scaccabarozzi, Pierre Râle, Fabrice Oehler, Aristide Lemaître, Laurent Lombez, Jean-Francois Guillemoles, Maria Tchernycheva, Jean-Christophe Harmand, et al.

To cite this version:

Hung-Ling Chen, Chalermchai Himwas, Andrea Scaccabarozzi, Pierre Râle, Fabrice Oehler, et al.. Cathodoluminescence mapping for the determination of n-type doping in single GaAs nanowires. IEEE PVSEC, Jun 2017, Washington, United States. hal-02181001

HAL Id: hal-02181001 https://hal.archives-ouvertes.fr/hal-02181001 Submitted on 11 Jul 2019

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Cathodoluminescence mapping for the determination of n-type doping in single GaAs nanowires Hung-Ling Chen1, Chalermchai Himwas1, Andrea Scaccabarozzi1,2, Pierre Rale1, Fabrice Oehler1, Aristide Lemaître1, Laurent Lombez2,3, Jean-François Guillemoles2,3, Maria Tchernycheva1, Jean-Christophe Harmand1, Andrea Cattoni1, Stéphane Collin1,2 1Centre of Nanoscience et de Nanotechnology, CNRS, University Paris-Sud/Paris-Saclay, Marcoussis, France 2Institut Photovoltaïque d'Ile-de-France (IPVF), Antony, France 3Institut de Recherche et Développement sur l’Energie Photovoltaïque (IRDEP) EDF/CNRS/Chimie Paris Tech, Chatou, France

Abstract — We present a new method to determine the doping In this work, we show that cathodoluminescence (CL) can level of n-type semiconductors at the nanoscale. Low- be used to assess the doping level of n-type GaAs at the temperature and room-temperature cathodoluminescence (CL) measurements are carried out on single Si-doped GaAs nanoscale. CL maps of single GaAs nanowires are measured nanowires. The spectral shift and the broadening of luminescence at 20 K and room temperature. They show a high homogeneity spectra are a signature of an increased density of electrons. They along the wire, and the peaks of CL spectra shifting to higher are compared to CL spectra of well-calibrated planar Si-doped energy (Burstein-Moss shift) is a signature of increased GaAs layers whose doping levels are determined by Hall electron concentration. We compare CL spectra of single measurements and compared to previous experimental studies. We infer a n-type doping of 1×1018cm-3 to 2×1018cm-3, with a high nanowires with CL spectra of planar Si-doped GaAs with spatial homogeneity along the nanowire. These results show that well-calibrated doping levels, and with data published in cathodoluminescence provides an alternative way to probe previous studies. We infer electron concentrations in GaAs carrier concentration in nanostructured and polycrystalline nanowires of 1-2×1018cm-3. semiconductors, and to map the spatial inhomogeneity of dopants. II. -EXPERIMENTAL

I. INTRODUCTION GaAs nanowires were grown by molecular beam epitaxy (MBE) on Si(111) substrate. Un-intentionally doped GaAs Semiconductor nanowires provide a new route toward low nanowires were firstly grown by self-catalyzed vapor-liquid- cost photovoltaics and multi-junction architectures. They solid method. Ga catalyst droplet was then crystallized by present natural light-trapping properties. Their small lateral exposing the sample to As flux only. Finally, Si-doped GaAs dimension allows lattice mismatch growth and enables direct shell was grown using standard conditions similar to planar integration of III-V nanowires on silicon cells, or to fabricate films. Nanowire length was measured to be 3-4 μm and flexible devices [1,2]. However, controlling the doping is a diameter about 200 nm. Nanowires are dispersed on a Si major issue in the fabrication of nanowire-based solar cells. It substrate for CL measurements. is especially crucial for core-shell structures due to small Cathodoluminescence (CL) measurements were performed radial dimension, and numerical calculations have shown that with an Attolight “Chronos” quantitative the requirement of doping level is higher for radial junction cathodoluminescence microscope. The mean electron current than axial junction solar cells [3,4]. is of the order of 1nA and the acceleration voltage is 6kV. CL Hall measurement is the conventional method used for the spectra were recorded on an Andor Newton CCD camera with characterization of doping in planar semiconductor layers. For a Horiba dispersive spectrometer (grating: 150 grooves/mm). nanowires, electrical methods can still be applied [5,6]. Light is collected through an achromatic reflective objective However, contacting a single nanowire and forming good with a numerical aperture of 0.72. Luminescence spectra are ohmic contacts requires considerable technical efforts and can corrected for the optical response of the collection and hardly be applied to a large number of nanowires. Contactless detection system. optical methods based on terahertz spectroscopy and photoluminescence have shown usefulness in measuring the doping level of ensembles of nanowires [7,8], but they cannot III. RESULTS be used to assess the homogeneity of semiconductor A. Cathodoluminescence mapping nanostructures and for the characterization of single nanowires.

Fig. 1. Cathodoluminescence mapping of a single GaAs nanowire. (a) SEM image of the nanowire. (b) CL map of the total luminescence intensity integrated over the whole emission spectral range. (c) CL emission spectra along the nanowire showing a very homogeneous emission except close to the tip, where the Ga catalyst droplet was crystallized. Doping analysis is done in the homogeneous part of the nanowire. (d) Map of the peak energy of luminescence emission. (e) Map of the full width at half maximum (FWHM) of the CL peaks.

Figure 1 shows cathodoluminescence measurements of a doping. At lower energies, the luminescence tails also grow single GaAs nanowire dispersed on a Si substrate. The CL with increased carrier concentration as a result of Coulombic intensity is constant along the major part of the nanowire, potential from ionized donor and electron-electron interaction except at the nanowire tip (top of the images in Figure 1). This [10]. Both effects contribute to the broadening of the CL peak part has been grown by consuming the liquid Ga catalyst with increased doping. droplet, and changing the contact angle of liquid droplet typically produces different crystal phases (zinc-blende or wurtzite) [9]. A broadening of CL spectra is observed in this region (Figure 1c). In the following, we will only analyze CL spectra from the homogeneous region. Figures 1(d) and (e) show the maps of the luminescence peak energy and full width at half mawimum (FWHM).

B. Low-temperature and room-temperature CL spectra Low-temperature and room-temperature CL spectra were measured on several GaAs nanowires and Si-doped GaAs planar layers with various doping levels, see Figures 2 and 3. Fig. 2. Low-temperature (20K) cathodoluminescence spectra of The spectra present a single peak across the band gap of planar GaAs layers with various n-type doping concentrations (un- GaAs (1.424 eV at room temperature and 1.519 eV at very doped and Si-doped), and Si-doped GaAs nanowire. low temperature, 1.515 eV corresponds to exciton recombination for low doping samples). We observe a blue Luminescence broadening could also be due to shift of the peak energy with increasing electron inhomogeneity and carrier heating. The homogeneity is concentrations (Burstein-Moss shift). It originates from the ensured by high resolution CL mapping, as shown in Figure 1. electron filling in the conduction band. As the electron Fermi The effect of carrier heating was controlled by level rises above the conduction band minimum, the optical photoluminescence (PL) measurements performed on planar absorption as well as spontaneous emission spectrum shifts to layers and single nanowires at various excitation intensities higher energies compared to the nominal band gap values. For (not shown). PL spectra are in very good agreement with CL many III-V semiconductors, the conduction band has much spectra, except in the long wavelength range were slight smaller effective density of states than the valence band. carrier heating is visible in CL spectra. Its contribution is Hence the Burstein-Moss shift is typically observed for n-type negligible with respect to the total peak broadening.

horizontal dashed line indicates the band-gap of un-doped GaAs emission spectra GaAs at room temperature. 1 Emax=1.424 fwhm=0.026eV (un-doped) 0.9 Emax=1.427 fwhm=0.043eV (n=5.3×1017cm-3) 18 -3 0.8 Emax=1.433 fwhm=0.070eV (n=1.7×10 cm ) Emax=1.449 fwhm=0.088eV (nanowire) 0.7 Emax=1.456 fwhm=0.135eV (n=5.8×1018cm-3) 0.6

0.5

0.4

0.3 Normalized CL intensity CL Normalized 0.2

0.1

0 1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8 Photon energy (eV) Fig. 3. Room-temperature cathodoluminescence spectra of planar GaAs layers with various n-type doping concentrations (un-doped and Si-doped), and Si-doped GaAs nanowire.

C. Relation between CL spectra characteristics and electron concentration The two main characteristics of luminescence spectra are the energy position of the maximum peak (peak energy) and the full width at half maximum (FWHM). Figure 4 presents the Fig. 5. FWHM as a function of electron concentration at relation between the peak energy and electron concentration room temperature. Our CL measurements are compared with from various experimental data, and Figure 5 shows the experimental data available in the literature: Casey and Kaiser relation between FWHM and electron concentration (room [11], Neave et al. [12] and Szmyd et al. [13]. The horizontal temperature). line indicates the room temperature thermal energy.

Our measurements are very close to that of Casey and Kaiser [11], while other data show significantly larger blue shift and broadening. Presence of acceptor states related to the growth technique may contribute to dominant radiative recombination, leading to different luminescence features. Casey and Kaiser showed very little dispersion of CL FWHM with electron concentration, so the use of FWHM to determine electron concentration seems simple and accurate [11]. The peak energy showed larger variation, especially at high doping level when the spectra tend to flatten. Therefore, we mainly focus on the use of FWHM for the determination of the n-type doping level, and the peak energy is used as a control.

At low temperature, Neave et al. observed transitions due to acceptor states for MBE-grown Si-doped GaAs samples [12]. In our Si-doped samples, we see no evidence of emission in the lower energy part of the CL spectra (Figure 2). De-Sheng et al. investigated the impact of the low-temperature FWHM with electron concentration n and found a dependence in n2/3 [15]. We fit the low-temperature FWHM as a power function of n (Figure 6) and find the power factor very close to 2/3, Fig. 4. Peak energy as a function of electron concentration at which can be explained by the electron Fermi level above the room temperature. Our CL measurements are compared with experimental data available in the literature: Casey and Kaiser parabolic portion of the conduction band [15]. for CL measurements on melt-grown Te-doped GaAs [11], Neave et al. for PL Si-doped GaAs grown by MBE [12], Szmyd et al. for PL Se-doped GaAs grown by MOCVD [13] and Lee et al. for PL Si-doped GaAs grown by MBE [14]. The

The measured FWHM can now be used to determine the electron concentration of nanowires. We checked that there is no clear unexpected luminescence peak and the position of the maximal intensity is blue shifted. Figure 7 reports the room- temperature FWHM of CL spectra measured on planar Si- doped GaAs layers (red). They are used as a reference to determine the electron concentration of different nanowires of the same sample based on the FWHM of the emission peak (blue marks). Doping levels of Si-doped nanowires are of the order of 1×1018cm-3 to 2×1018cm-3. Using the spectral shape of luminescence, we then provide a robust way to assess the n- type doping level above degeneration threshold (~5×1017cm-3). More accurate estimation can be achieved by fitting the whole spectrum with physical model [16,17].

V. CONCLUSION In this work, high-resolution cathodoluminescence mapping of single GaAs nanowires has been performed at low (20K)

Fig. 6. FWHM as a function of electron concentration at low and room temperature. Cathodoluminescence spectra were temperature (20K). Our CL measurements are compared with compared to planar Si-doped GaAs layers whose doping levels experimental data from De-Sheng et al. for PL MBE GaAs:Te are determined by Hall measurements. The spectral shift and at 1.8K [15]. the broadening of luminescence spectra are a signature of an increased density of electrons. They are compared to experimental data available in the literature. We infer n-type 18 -3 18 -3 D. Determination of n-type doping in single GaAs nanowires doping of 1×10 cm to 2×10 cm , with a high spatial homogeneity along the nanowire. This method could be extended p-type nanowires, and polycrystalline thin films developed for photovoltaic applications. It provides an alternative way to probe carrier concentration in nanostructured and polycrystalline semiconductors, and to map the spatial inhomogeneity of dopants.

ACKNOWLEDGMENTS This work was partly supported by the French ANR projects NANOCELL (ANR-RF-2015-01) and HETONAN. The Attolight cathodoluminescence tool was funded by public grants supported by the Region Ile-de-France in the framework of C’Nano IdF (nanoscience competence center of Paris Region), by the European Union (FEDER 2007-2013), and by the Labex GANEX (ANR-11-LABX-0014) and NanoSaclay (ANR-10-LABX-0035) as part of the "Investissements d'Avenir" program managed by the French National Research Agency (ANR).

Fig. 7. FWHM of room temperature CL spectra of planar GaAs layers as a function of their electron concentration determined by Hall measurements (red marks). FWHM of CL REFERENCES spectra measured on single GaAs nanowires are compared to planar layers in order to assess the n-type doping level ~1- [1] Gaute Otnes, Magnus T. Borgström, "Towards high efficiency 2×1018cm-3 (blue circles and error bars). nanowire solar cells," Nano Today, in press (DOI: 10.1016/j.nantod.2016.10.007), 2017.

[2] M. Yao et al., "Tandem Solar Cells Using GaAs Nanowires on [10] E. O. Kane, “Band tails in semiconductors,” Solid-State Si: Design, Fabrication, and Observation of Voltage Addition," Electronics, vol. 28, no. 1, pp. 3–10, Jan. 1985. NanoLetters, vol. 15, no. 11, pp. 7217–7224, 2015. [11] H. C. Casey and R. H. Kaiser, “Analysis of N Type GaAs with [3] R. R. LaPierre, “Numerical model of current-voltage Electron Beam Excited Radiative Recombination,” J. characteristics and efficiency of GaAs nanowire solar cells,” Electrochem. Soc., vol. 114, no. 2, pp. 149–153, Feb. 1967. Journal of Applied Physics, vol. 109, no. 3, p. 034311, Feb. [12] J. H. Neave, P. J. Dobson, J. J. Harris, P. Dawson, and B. A. 2011. Joyce, “Silicon doping of MBE-grown GaAs films,” Appl. Phys. [4] M. Yao et al., “GaAs Nanowire Array Solar Cells with Axial p– A, vol. 32, no. 4, pp. 195–200, Dec. 1983. i–n Junctions,” Nano Lett., vol. 14, no. 6, pp. 3293–3303, juin [13] M. Szmyd, P. Porro, A. Majerfeld, and S. Lagomarsino, 2014. “Heavily doped GaAs:Se. I. Photoluminescence determination [5] K. Storm et al., “Spatially resolved Hall effect measurement in a of the electron effective mass,” Journal of Applied Physics, vol. single semiconductor nanowire,” Nat Nano, vol. 7, no. 11, pp. 68, no. 5, pp. 2367–2375, Sep. 1990. 718–722, Nov. 2012. [14] N.-Y. Lee et al., “Determination of conduction band tail and [6] O. Hultin, G. Otnes, M. T. Borgström, M. Björk, L. Samuelson, Fermi energy of heavily Si-doped GaAs by room-temperature and K. Storm, “Comparing Hall Effect and Field Effect photoluminescence,” Journal of Applied Physics, vol. 78, no. 5, Measurements on the Same Single Nanowire,” Nano Lett., vol. pp. 3367–3370, Sep. 1995. 16, no. 1, pp. 205–211, Jan. 2016. [15] De-Sheng, Y. Makita, K. Ploog, and H. J. Queisser, “Electrical [7] S. Arab, M. Yao, C. Zhou, P. D. Dapkus, and S. B. Cronin, properties and photoluminescence of Te-doped GaAs grown by “Doping concentration dependence of the photoluminescence molecular beam epitaxy,” Journal of Applied Physics, vol. 53, spectra of n-type GaAs nanowires,” Applied Physics Letters, no. 2, pp. 999–1006, Feb. 1982. vol. 108, no. 18, p. 182106, May 2016. [16] Wurfel, “The chemical potential of radiation,” J. Phys. C: Solid [8] H. J. Joyce, J. L. Boland, C. L. Davies, S. A. Baig, and M. B. State Phys., vol. 15, no. 18, p. 3967, Jun. 1982. Johnston, “A review of the electrical properties of [17] J. K. Katahara and H. W. Hillhouse, "Quasi-Fermi level splitting semiconductor nanowires: insights gained from terahertz and sub-bandgap absorptivity from semiconductor conductivity spectroscopy,” Semicond. Sci. Technol., vol. 31, photoluminescence," J. Appl. Phys., vol. 116, no. 17, p. 173504, no. 10, p. 103003, 2016. 2014. [9] F. Glas, J.-C. Harmand, and G. Patriarche, “Why Does Wurtzite Form in Nanowires of III-V Zinc Blende Semiconductors?,” Phys. Rev. Lett., vol. 99, no. 14, p. 146101, Oct. 2007.